Latent Variable Mixture models to track Longitudinal Differentiation Patterns
In everyday educational work or clinical practice, development is something that is aspired, monitored, and worked on to make sure that students learn or patients improve upon their current condition.
What happens if the ruler changes?
Usually, progression addresses the question how far you moved along a ruler and assumes that as long as the same ruler (i.e., measurement instrument) is used, scores can naturally be compared across time. If the ruler would change or what the ruler is trying to measure changed throughout the process, the common ground for comparisons disappears. Hence, you risk comparing apples and oranges.
Yet, in some situations, you are unable to use the same ruler or have to redefine what you are measuring, and changing from an apple into an orange would be an actual sign of development!
Redefinition might be necessary
For instance, we can expect that progressing from the level of a novice student-teacher to the level of a more expert veteran teacher is not simply "growing" more of the same competence, but that instead it actually requires redefining your understanding and definition of the teaching practice. Similarly, the reported quality-of-life of patients might also undergo a response shift as they redefine/re-evaluate what quality of life means for them as a person while disease progresses or impactful events such as operations happen.
In such situations, we need to rethink to what extent simple growth comparisons remain useful and how to provide alternative ways to measure and model such differentiating developmental patterns. This project aims to develop sound statistical procedures to accommodate the tracking of development in such settings.
From a psychometric perspective, the project will focus on longitudinal measurement equivalence and on the creative use of latent variable mixture models to account for inequivalent progress trajectories and individual differences in development.
We will make R-packages and Shiny-applets available here for methods and statistical models developed in the project, and R-code for papers written in the context of the project.
The project is funded by the Research Council of Norway